“Popular Support for Regional Economic Integration in Latin America,” Journal of Latin American Studies, Volume 31, Part I, February,
1999, pp. 129-150.
J. Lat. Amer. Stud. , –
Printed in the United Kingdom
#  Cambridge University Press

Popular Support for Regional
Economic Integration in Latin
America
M I T C H E L L A. S E L I G S ON
Abstract. Regional economic integration schemes abound in Latin America, yet
very little is known about the degree of popular support for such programmes.
Now that democratic regimes rule in almost all of Latin America, public opinion
can have an important impact on national policies. This paper examines the
attitudes toward economic integration in  mainland Latin American countries
with interview data from over , people. Bi-variate and multivariate analyses
are conducted on the factors related to support for regional integration.
Perceived benefits of integration and perceptions of personal and national
economic situation prove to be important factors, as are higher levels of
education, support for democracy and gender (males favour integration more
than females). Two novel findings are that a positive opinion of the European
Union along with satisfaction with the functioning of democracy are both linked
to greater support for integration.
Although the world’s most successful large-scale regional economic
integration scheme has arisen in Europe, since the end of World War II
Latin American nations have experimented with numerous integration
programmes. Success in Latin America, however, has been far more
limited, and once-promising schemes like the Central American Common
Market fell into disarray as a result of international conflicts and domestic
upheavals." Nonetheless, the recent rapid expansion of Mercosur, the
trade group currently dominant in the ‘ Southern Cone ’ of South
America, and the revitalisation of both the Andean Pact and the once
moribund Central American Common Market has led some to suggest
optimistically that Latin America might now be following on Europe’s
integrationist footsteps. Their hope is that many of the economic gains
that have accrued to Europe as a result of economic integration could also
be won by Latin America if it were to develop strong regional or subregional trading blocs.
What factors might promote (or hinder) support for integration among
the governments of Latin America ? For the first thirty years after World
Mitchell A. Seligson holds the Daniel H. Wallace Chair of Political Science at the
University of Pittsburgh.
" For a review of the Central American case and recent efforts to reactivate the regional
economic group see Victor Bulmer-Thomas (ed.), IntegracioT n regional en CentroameT rica
(San Jose! , ).

Mitchell A. Seligson
War II, integrationist efforts in Latin America were promoted almost
exclusively by small groups of technocrats and economic elites, who were
largely able to operate without regard for potential public (especially
labour) opposition since at that time most of Latin America was ruled by
authoritarian regimes. By the late s, however, Latin America had
abandoned (at least for the moment) its authoritarian past. Today, with
the exception of Cuba, all of Latin America is governed by elected
regimes, even if the depth and stability of democracy in a number of
countries remains an open question.
This paper examines popular support for and opposition to regional
economic integration in Latin America. The probability that Latin
American governments will move forward or turn their backs on the
integrationist project depends today far more on public opinion than it
did in the authoritarian past. In democracies, public opinion can count a
great deal, and strong opposition to integration might well spell the end
of the regionalist movement in Latin America. Labour unions, for
example, can oppose integrationist measures, as many did in the United
States when NAFTA was being debated. Popular opposition to NAFTA
was one of the main arguments used by Congressional opponents to vote
against the treaty, and nearly led to its defeat. Public opinion, once
thought to be relevant in democracies only on matters of domestic
politics, has now been shown by extensive academic research to have a
large impact on key foreign policy issues.# Moreover, public opinion on
foreign policy issues, once believed to be almost random in nature, has
now been shown to be cohesively structured. Such structure has also been
shown to exist in Latin America.$
In the case of European integration, an extensive body of research has
been developed on the importance of public opinion for the furtherance
of the programme.% It was once believed that mass attitudes were largely
irrelevant to the complex technical issues of European integration, but the
 rejection by Danish voters of the Maastricht treaty showed that
# For a good literature review of the linkage between public opinion and public policy
see Lawrence R. Jacobs and Robert Y. Shapiro, ‘ Studying Substantive Democracy ’,
PS : Political Science & Politics, vol. , no.  (March ). For a study that shows the
direct effect of US public opinion on foreign policy, see Bruce Russett, Thomas
Hartley and Shoon Murray, ‘ The End of the Cold War, Attitude Change, and the
Politics of Defense Spending, ’ PS : Politics & Society, vol. , no. (March ).
$ Jon Hurwitz, Mark Peffley and Mitchell A. Seligson, ‘ Foreign Policy Belief Systems in
Comparative Perspective : The United States and Costa Rica ’, International Studies
Quarterly, vol. , no.  (September ).
% Robert J. Shepherd, Public Opinion and European Integration (Westmead, ) ; Ronald
Inglehart, ‘ Long Term Trends in Mass Support for European Unification ’, Government
and Opposition, vol. , no.  (Summer, ), pp. – ; Ronald Inglehart, Culture
Shift in Advanced Industrial Society (Princeton, ).
Regional economic integration in Latin America

without public support, progress on the integration process could be
suspended or even reversed. According to an extensive study of public
opinion data collected throughout the EU for over ten years, Richard
Eichenberg and Russell J. Dalton have concluded, ‘ When seen in the
context of the post-Maastricht debate in Europe, our results suggest that
the link between domestic citizen consensus and the content of
international agreements will become even closer. ’&
While European opinions on the EU have been extensively studied, far
less is known, however, about other regional integration schemes. With
the exception of NAFTA, the role of public opinion in regional economic
schemes in Latin America has been largely devoid of study. The only
published research has covered individual countries or small groupings,
such as integration in Central America.' Very recently, however, a new
data source has been made available that allows us to take a careful look
at the way that Latin Americans think about economic integration. The
Latinbarometer, modelled after the Eurobarometer on which so much
integration research has been based, covers all of the mainland countries
from Mexico to the tip of South America, with the exception of Belize,
Surinam and the Guyanas.( The  Latinbarometer involved over
, interviews in  countries. The countries covered and sample sizes
are shown in Table . In most countries the samples hovered around ,
respondents, except in Venezuela where the number was , and in
Bolivia and Paraguay where the samples were smaller.) These are each
& Richard Eichenberg and Russell J. Dalton, ‘ Europeans and the European Community :
The Dynamics of Public Support for European Integration ’, International Organization,
vol.  (Autumn ), p. .
' Mitchell A. Seligson and Ricardo Co! rdova. ‘ Integration and Disintegration in Central
America : From – ’, in Lawrence A. Herzog (ed.), Changing Boundaries in the
Americas : New Perspectives on the US-Mexican, Central American, and South American
Borders (San Diego, ), pp. – ; and Mitchell A. Seligson ‘ Actitudes de la
Poblacio! n Centroamericana Frente a la Integracio! n Polı! tica y Econo! mica. ’ Anuario de
Estudios Centroamericanos, vol. , no.  (), pp. –. For a comparison of mass and
elite attitudes see Mitchell A. Seligson and Marı! a Pı! a Scarfo, ‘ El pu! blico y los
legisladores : apoyo para la integracio! n regional centroamericana ’, in Victor BulmerThomas, (ed.), CentroameT rica en reestructuracioT n : IntegracioT n regional en CentroameT rica (San
Jose! , ), pp. –.
( The data from the  Latinbarometer were made available by the European Union,
via the Centro de Investigacio! n, Promocio! n y Cooperacio! n Internacional of Spain and
the United Nations Development Programme in Central America.
) In the published summary of the Central American cases a slight variation of two to
three cases was found for Costa Rica, Nicaragua and Panama from the data set reported
on here. The variation is a result of ambiguous coding of the country location for a
total of eight interviews out of the more than , in the data base. See PNUD
(Programa de las Naciones Unidas para el Desarrollo), Desarrollo Humano Sostenible,
Informe LatinobaroT metro : Consolidado de CentroameT rica, Proyecto CAM.. (San Jose! ,
Costa Rica, ). For a general discussion of some of the survey data for Central
America see Proyecto Regional de Gobernabilidad para Centroa! me! rica, El desafıT o

Mitchell A. Seligson
Table . Sample distribution by country
Country
Sample
Per cent
 Argentina
 Bolivia
 Brazil
 Colombia
 Costa Rica
 Chile
 Ecuador
 El Salvador
 Guatemala
 Honduras
 Mexico
 Nicaragua
 Panama
 Paraguay
 Peru
 Uruguay
 Venezuela
Total
,

,
,
,
,
,
,
,
,
,
,
,

,
,
,
,
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
national probability samples. For the data set as a whole,  per cent of
the respondents were male, and  per cent female.
In the analysis that follows, the data for the  countries have been
weighted to reflect the differing population sizes of the countries from
which the samples have been drawn, and then pooled so that an overall
analysis of trends in Latin America can be examined.* Analyses of
European data on integration have taken the same approach, pooling the
data for all of the countries, or looking at the original members of
European Community and then the expanded (current) membership of
the EU. Since there is no Latin America-wide market in existence at the
moment, another approach to this data set would be to compare regional
groupings with Latin America (e.g., Andean Pact, Mercosur, the Central
American Common Market, etc.), while an even more micro-approach
would be to look at each of the  countries in isolation. That, however,
is not the approach taken in this paper, which seeks to find overall trends
for Latin America as a whole. In the future, researchers may wish to focus
on differences among the trading pacts and\or differences among
countries. It is important to note at the outset, however, that support for
democraT tico : Reflexiones de las sociedades centroamericanas ante el resultado del LatinobaroT metro
‰‰†, PNUD, San Jose! , Costa Rica,  ; and Marta Lagos, ‘ Latin America’s Smiling
Mask ’, Journal of Democracy, vol. , no.  (July ), pp. –.
* The population data are mid- estimates from the World Bank, World Development
Report, ‰ˆ‰ (Washington, D.C., ), pp. –.
Regional economic integration in Latin America

integration in Latin America is not a function of the relative level of
economic development of the countries involved. Inclusion of GDP data
(as shown below), makes no difference in the overall patterns uncovered.
The present study follows the analytical tradition of the analysis of
public opinion data on the EU, as developed by the ‘ Beliefs in
Government ’ programme of the European Science Foundation. A 
Oxford University Press volume summarises what is known about public
opinion for the European case."! The analysis in this paper begins with
opinions and ‘ non-opinions ’, that is, the extent to which Latin Americans
hold views on economic integration. It then goes on to examine
support\opposition to integration, and the perception of benefits received
to date of the various regional trading pacts. The data analysis then
focuses on the factors that relate to higher or lower support for
integration among Latin Americans, and concludes with a multi-variate
analysis of those factors.
Opinions and Non-Opinions
Before looking at the opinions about integration in Latin America, it is
first necessary to separate out those who have an opinion on the subject
from those who do not. If, for example, few Latin Americans had any
opinion on the subject, it would misrepresent the overall picture to
present the percentages from the small minority that did have an opinion.
We can assess the absence of opinions on Latin American integration by
utilizing one of the two key questions asked in the Latinbarometer survey.
All respondents were asked : ‘ In general, are you in favour or against
economic integration of the countries of Latin America, even when this
might imply some costs or sacrifices for [fill in country of survey] ’. This item
might seem skewed, since it emphasises the costs of integration without
noting the benefits. Prior focus group research has shown, however, that
unless the costs are mentioned, respondents are too quick to respond
positively to questions on economic integration without thinking through
all of the implications. This would be the equivalent of asking a
respondent, ‘ Would you like to earn more money ? ’ without noting that
to do so would require additional hours of work per week. When the
possible costs involved are noted, a more realistic estimate of support is
obtained. The percentage of respondents in each country who did not
have an opinion on this item are shown in Figure . For the sample as a
whole, . per cent of the respondents did not have an opinion (pro or
con) on Latin American economic integration. As the Figure shows,
"! Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized
Governance, Beliefs in Government, Volume Two (Oxford, ).

Mitchell A. Seligson
50
With no opinion (%)
40
41
35
30
30 29
20
24
22 21
21
19
16 16 16
10
15 14 14
10
9
G
ua
t
H em
on al
Co du a
s r
El ta as
Sa Ri
c
Ve lva a
ne do
zu r
el
Pe a
ru
A Ch
rg i
e le
Pa nti
ra na
gu
a
B y
U ra
ru zi
g l
M uay
ex
Ec ic
ua o
Pa dor
N n
ic am
ar
ag a
Bo ua
Co liv
lo ia
m
bi
a
0
Fig. . Non-opinion on economic integration in Latin America.
Table . Predictors of non-opinion : logistic regression results
Variable
Attentiveness to political news
Years of education
Wealth
Sexa
Ageb
Constant
a
b
B
k.
k.
k.
.

.
..
.
.
.
.
.
.
d.f.






P
.
.
.
.
.
.
R
k.
k.
k.
.

Sex coded :  l male ;  l female.
Wealth measured by a -item index of ownership of various household appliances.
however, in Central America, with the exception of Panama and
Nicaragua, non-opinion is as high as one-third to two-fifths of the
samples.
How do those who hold opinions differ from those without opinions
on the issue of regional integration in Latin America ? Logistic regression
analysis was utilised on the pooled survey comparing opinion holders to
those with non-opinion. The results are shown in Table . As has
frequently been found in European surveys on integration, the main
determinants of non-opinion in Latin America are socio-economic and
demographic factors. Specifically, those who expressed no opinion in
favour or against Latin American economic integration were significantly
more likely to be older, poorer, less well educated, less attentive to
political news and female.
Regional economic integration in Latin America

50
With no opinion (%)
40
30
20
Mean
10
G
ua
t
H em
on a
Co du la
s r
El ta as
Sa Ri
c
Ve lvad a
ne o
zu r
el
Pe a
ru
A Ch
rg il
e e
Pa nt
ra ina
gu
ay
U Bra
ru zi
gu l
M ay
e
Ec xic
ua o
Pa do
N na r
ic m
ar a
ag
Bo ua
Co li
lo via
m
bi
a
0
Fig. . Non-opinion on economic integration in Latin America : by Gender :
, male ; , female.
There is some minor variation in the pattern of non-response by
country. For example, in Figure  the responses have been broken down
by gender. In every country except Guatemala, females are significantly
less likely to hold an opinion than males on this subject. In Europe the
gender gap exists as well, but it is far narrower, averaging less than  per
cent.""
The other socio-economic and demographic variables shown in the
regression analysis are significant predictors in each country, although the
strength of association varies from country-to-country. For example,
Figure  shows that in every country in the survey, those who had
an opinion on economic integration had a higher average level of
education (measured by years in school), than those who did not have an
opinion.
The Latinbarometer utilised a second question to assess opinion on
regional integration by asking, ‘ Taking everything into account, would
you say that [fill in country] has benefited a lot, sufficiently, a little, not at
all, by belonging to the [Andean Pact, Mercosur, etc.] ’. The analysis of the
responses to that item forms a component of the survey results reported
"" Bernhard Wessels, ‘ Development of Support : Diffusion or Demographic Replacement ? ’ in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and
International Governance (Oxford, ).

Mitchell A. Seligson
14
13
12
11
Years of education
10
9
8
7
6
5
4
3
2
1
A
rg
en
t
Bo ina
liv
ia
Co Bra
z
l
Co om il
st b
a ia
Ri
ca
C
Ec hi
El ua le
Sa do
G lva r
ua d
te or
H ma
on l
du a
M ra
N ex s
ic i
ar co
a
Pa gu
a
n
Pa am
ra a
gu
ay
U Per
r
u
Ve ugu
ne ay
zu
el
a
0
Fig. . Non-opinion on economic integration in Latin America : by education :
, opinion ; , non-opinion.
below, but non-response patterns and correlates are similar for the
support–oppose integration item analysed above, so they will not be
repeated here."# Overall, then, the patterns of non-opinion on economic
integration are similar throughout Latin America, and are a predictable
function of socio-economic and demographic characteristics.
Support for and benefits of regional integration in Latin America
Support for integration
This paper has shown that on average, more than four-out-of-five Latin
Americans have an opinion about the desirability of economic integration.
For those who did have an opinion in the  wave of surveys, how
much support was there among Latin Americans for economic
"# Actually, there is yet a third series measuring opinion–non-opinion in the
Latinbarometer. This series, however, asks respondents about each specific regional
trading bloc, namely, Mercosur, NAFTA and the Pacto Andino. Analysis of those
questions shows, not surprisingly, that knowledge is far higher of the specific pact for
the samples drawn from the countries which are members of that pact. Unfortunately,
no question was asked on the Central American Common Market, thus excluding those
six countries from the analysis. In order to avoid having to present results for each
trading block and for each set of countries belong to that bloc, these variables are not
analysed here.
Regional economic integration in Latin America

100
90
85
Favouring integration (%)
80
Mean
70
72
69 69
60
50
57
76 77
74 75 76
88
92
80 81
61 63
54
40
30
20
10
M
e
Pa xico
G na
ua m
te a
H ma
on l
El d a
Sa ura
Ve lvad s
n o
Co ezu r
st el
a a
Ri
c
N Br a
ic az
ar il
ag
u
Ch a
U
ru ile
Ec gua
u y
Pa ad
ra or
gu
Bo ay
liv
ia
A Pe
rg r
en u
Co ti
lo na
m
bi
a
0
Fig. . Support for integration in Latin America (includes only those who
hold an opinion).
integration ? To answer this question, the ‘ non-opinions ’ are first
excluded from the data base so that the focus from this point on is on
those who hold an opinion. The answer to the question is given in
Figure . In all but four countries, two-thirds of the sample express
support, and in nine out of the  countries, three-quarters or more of
those with an opinion are in favour of integration. The pattern that
emerges shows that support is weakest in Mexico and Central America,
and strongest in South America, especially in Paraguay, Bolivia, Peru,
Argentina and Colombia.
Evidentially, there is a ‘ north–south ’ dimension in determining
support for regional integration in Latin America. It may well be that
Mexicans and Central Americas are more interested in integrating with
North America than they are with South America. Unfortunately, the
Latinbarometer questionnaire does not examine NAFTA versus Latin
American integration programmes, so it is not possible to probe this
finding further with the data set. It is suggestive, however, that even
though Chile has been designated as the next country to enter NAFTA
(pending, of course, approval by all current members, a step which is by
no means assured at the time of writing), Chileans support economic
integration with Latin America to a far greater extent than do Mexicans
and many Central Americans.

Mitchell A. Seligson
Table . Support–opposition to Latin American integration

















Argentina
Bolivia
Brazil
Colombia
Costa Rica
Chile
Ecuador
El Salvador
Guatemala
Honduras
Mexico
Nicaragua
Panama
Paraguay
Peru
Uruguay
Venezuela
Favour (%)
Oppose (%)
No opinion (%)
Total (%)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
The North–South split becomes even more evident when non-opinion
is factored into the analysis. Table  shows that for every country, when
non-opinion is included in the analysis, support for integration declines.
In the northern zone of Latin America, only in Nicaragua and Costa Rica
are there majorities in favour of integration when the full sample,
including those with no-opinion, is analysed. Even in those countries,
however, as noted above, those who favour integration outnumber those
who oppose it, and for South America, majorities favour regional
integration even when non-opinion is included.
Overall, then, policy makers in South America who wish to press for
regional economic integration can be confident that in every country
majorities of those holding an opinion would support it, and in most
countries, only less than a fifth of the population opposes economic
integration. Only in Mexico and Panama does more than one-third of the
population oppose regional integration.
Benefits of integration
Support for economic integration is important, of course, but citizens
want to see payoffs from policy. What is the perception in Latin America
of the benefits of regional integration to date ? To what extent do Latin
Americans believe that their current regional trading schemes (Andean
Pact, Mercosur, Central American Common Market, etc.), have been
beneficial ? Figure  presents a very different pattern from that obtained
from the analysis of support\opposition to economic integration. With
the exception of Colombians, who not only had the lowest level of non-
Regional economic integration in Latin America

100
90
Degree of benefit
80
76
70
60
Mean
50
40
44 45
60 62
57 57 58 59 59
57
56
55
55
55
48 50
30
20
10
Pa
ra
gu
M ay
ex
ic
o
U Per
ru u
H gu
on ay
du
ra
C s
Ec hil
G ua e
ua d
t or
Co em
st ala
Ve a R
ne ica
zu
el
Br a
a
Pa z
na il
m
Bo a
A li
rg v
El en ia
Sa tin
a
N lva
ic do
ar r
a
Co gu
lo a
m
bi
a
0
Fig. . Benefits from integration (among opinion holders). Scale : none l  ; a little l
 ; sufficiently l  ; a lot l . Includes only those who hold an opinion.
opinion but also have the highest level of favourable opinion, the benefits
of regional integration received to date are seen, on average, as being only
modest. Most countries cluster very closely around the average for all
countries (. on the -point scale used to present this item), with only
four countries (Paraguay, Mexico, Peru and Uruguay) falling slightly into
the negative range. Moreover, although not shown in Figure , for the
sample as a whole, non-opinion is larger, moving from about one-fifth of
respondents on the question about support for integration to about onequarter of respondents on the benefits question.
Combining the information gleaned from the survey items analysed
thus far, two conclusions can be drawn. First, there is strong general
support for Latin American regional integration. Secondly, the benefits of
existing regional integration schemes are perceived as having been only
modest in all countries, except in Colombia, where benefits are perceived
as being greater. What explains variation in Latin America on support for
economic integration ? One might assume that it is largely a function of
the degree of industrial development of each country, since most of the
trade in Latin America generated by regional integration schemes has
emerged in the industrial sector. In fact, when national-level economic
development is examined as a predictor of support, no significant
association emerges."$ Moreover, there is no pattern linking exports (or
"$ Economic development was measured as the so-called ‘ real ’ GDP per capita for 
as calculated by the World Bank. Real GDP is based on purchasing power parity rather

Mitchell A. Seligson
exports per capita) and attitudes favourable toward regional economic
integration. In order to explain variation in levels of support, we must
look to individual-level factors, the analysis of which this paper now
turns.
Predictors of support for economic integration
‘ Pocketbook ’ issues
Research has shown that the main determinants of support for European
integration are ‘ pocketbook issues ’."% The strongest attitudinal predictor
of support for European integration is the extent to which respondents
feel that the integration movement has paid off economically for their
country."& In addition, those who feel that their national economy is doing
well, and that they personally are doing well economically, are more
supportive of integration. Even though the experience with economic
integration in Latin America is more recent and more limited than in the
European case, the individual-level patterns we find in this analysis mirror
very closely those uncovered in Europe.
The impact of the belief that one’s country has benefited from regional
trading pacts on support for Latin American integration is shown in
Figure . The relationship is obviously very strong, and holds equally well
for men and women. Among Latin Americans who believe that the
existing regional trade pacts have been of no benefit to their country,
fewer than two-fifths support Latin American economic integration,
whereas among those who believe that their country has benefited a great
deal, over  per cent support integration. Female support for integration
is virtually identical to that of the males in the sample, except at the very
than exchange rates. See World Bank, World Development Report, ‰‰‡ (Washington,
D.C., ).
"% For an extensive analysis see Richard Eichenberg and Russell J. Dalton, ‘ Europeans
and the European Community : The Dynamics of Public Support for European
Integration.’ Distinctions are made between so-called ‘ socio-tropic ’ issues and
‘ pocketbook ’ issues, the former being focused on the general state of the economy
while the latter looking at personal economic issues. Analysis of the Latinbarometer
data finds that both sets of economic issues have an impact on support for
integration, with only relatively small differences between them. Researchers
interested in this distinction could subdivide the items in the ‘ pocketbook ’ scale
used here.
"& See Charles Lyon, Jeffery J. Mondak and Anke Grosskopf, ‘ Economic Perceptions and
Mass Support for European Integration : Exploring the Dimensions of Sociotropic
Behavior ’, Paper presented at the Midwest Political Science Association Annual
Meetings (Chicago, Illinois, April ).
Regional economic integration in Latin America

100
Supporting integration (%)
90
80
70
60
50
40
30
Not at all
A little
Sufficiently
At lot
Benefits of regional pacts
Fig. . Support for regional integration and perceived benefits of regional pacts.
Gender : ——, male ; - - -, female.
highest levels of the benefit variable, where support among females is
slightly lower.
Not only do Latin Americans link their perception of the economic
benefits of regional pacts to their support for regional integration, but
they also link it to their perception of economic issues more broadly. The
survey asked a series of questions regarding respondent perception of
their national and their own personal economic situation. A direct, linear
relationship is found for variables measuring perception of the future of
the nation’s economy as well as the respondent’s own current and
anticipated economic situation. The relationship between these variables,
combined into an index, is shown in Figure ."' Although the results
show a clear and strong relationship between Latin Americans ’ perception
of their nation’s and their own economic situation and support for
"' The index is comprised of four variables : () perception of the current economic
situation of the country compared to a year previously ; () perception of the economic
situation one year in the future ; () perception of respondent’s family situation one year
in the future and () degree of fear over personal unemployment in years to come. The
overall index has an Alpha reliability of ., and produces a single factor in a principal
components analysis. The item least closely related to the other three is P (fear over
personal unemployment). When this item is dropped, the Alpha rises to .. In the
survey, these items are labels P, P, P and P. The responses for each item are
recoded here on a – basis to give them an equivalent range, and the sum divided
by . An analysis of the items comparing retrospective concerns (P) versus
prospective concerns (P, P and P) does not produce any substantive changes in
the results.

Mitchell A. Seligson
Favouring integration (%)
100
90
80
70
60
50
0
8 13 21 25 29 33 38 42 46 50 54 58 63 67 71 75 79 83 88 92 100
Index of perception of economic situtation
Fig. . Perception of the economy and support for integration.
integration, it is also clear from Figure  that even among those who are
most pessimistic about their economy, about two-thirds still support
integration. Nonetheless, these findings suggest quite clearly the identical
pattern found in Europe, namely, that support for integration is highly
contingent on pocketbook economic issues. In countries where the
economy is headed in a downward direction one must anticipate declining
support for integration. Of course, if the integration process itself
produces strong rewards, and the economies grow, it is also possible to
predict a reinforcement of support for integration.
Socio-economic and demographic factors
Immediate pocketbook issues clearly have their impact on support for
integration, but underlying socio-economic and demographic factors also
play a role. In Europe, those with higher education, professionals and
males are all more supportive of integration than those with lower
education, manual workers and females. Gender has already been noted as
playing a role in non-opinion in the Latin American data, and in this
section, this factor as well as socio-economic factors are analysed."(
In order to properly examine the role of gender on support for
economic integration it is necessary to control for education, since the
survey data show that males have significantly more education than
"( An examination of age found no relationship in the data set. In the analysis of
European data a number of studies have been conducted looking at age cohorts and
their changing views over time, but since the Latinbarometer does not yet offer data
sets over time, this analysis cannot be conducted here.
Regional economic integration in Latin America

Favouring integration (%)
90
80
70
60
50
Pr
im
ar
y
Ill
ite
ra
t
e
in
co
m
pl
Pr
et
im
e
ar
y
co
Se
m
co
pl
nd
et
e
ar
y
in
co
Se
m
co
pl
nd
et
e
ar
y
co
m
pl
et
So
e
m
e
un
iv
er
U
si
ni
ty
ve
rs
ity
de
gr
ee
40
Fig. . Gender and support for integration : controlled by education. Gender :
——, male ; - - -, female.
females, although the difference is not large in absolute terms. Figure 
shows the relationship between gender and support for Latin American
economic integration, controlled for education. Two findings are apparent
from this analysis. First, education plays a strong role in predicting
support for Latin American economic integration ; among those who are
illiterate, support hovers around  per cent, whereas about  per cent
of those who have completed university degrees support integration.
Secondly, the gender gap opens for those with secondary education and
beyond, but for illiterates, females are slightly more supportive than
males. Overall, however, the gender gap is not wide, with education
clearly explaining far more of the difference in support than gender.
Research in Europe has often found that professionals are more
supportive of integration than blue-collar workers. As a result, an
examination by occupation is conducted on our Latin American data base.
Figure  shows the results, which present no surprises. Once again we
see that for all occupational categories, mean support for integration is
positive. Yet, among those with the highest level professional occupations, support is nearly at the  per cent level, whereas it drops down
to only two-thirds among farmers.
The final socio-economic variable linked to support for integration is
wealth. As in many surveys, the Latinbarometer did not ask directly about
income since many respondents refuse to answer this question or provide
a deceptive answer. A better measure of the respondent’s economic

Mitchell A. Seligson
Favouring integration (%)
100
90
91
84
80
75
70
75
74
71
70
68
H
ig
hle
ve
In
de
le
pe
xe
nd
cu
en
tiv
tp
e
M
ro
id
fe
-le
ss
io
ve
ns
Sa
le
xe
la
rie
cu
d
tiv
pr
e
of
es
si
on
s
Em
pl
Bu
oy
si
ee
ne
In
ss
de
o
w
pe
ne
nd
r
en
tw
or
ke
r
Fa
rm
er
60
Fig. . Occupation and support for integration.
success is obtained by asking about the possession of capital goods,
especially home appliances and cars. The Latinbarometer asked about 
of these goods, and an index was created from them.") As expected,
wealth was positively associated with support for integration, but the
relationship was not strong (r l . ; Sig. .). Since wealth and
education are so tightly linked, it turns out that the relationship between
wealth and support for integration becomes spurious when controlled for
education. The multivariate analysis presented at the conclusion of this
paper excludes wealth because it becomes an insignificant predictor when
education is entered into the equation.
The ‘ role model ’ of the EU
Relations between Latin America and Europe vary widely from countryto-country. Generally speaking, Southern Cone countries have paid closer
attention to Europe than have the countries in Central America. In recent
") Although the capital goods measured in the survey vary widely in cost, from a
microwave to an automobile, no attempt was made to weight these items in the scale.
Rather, a summative scale was utilised, but since there was some missing data, mean
scores of the individual respondent were assigned to each person who answered six or
more of the  items. For those who had a larger proportion of missing data, the case
was coded as missing. The goods were : colour TV, video player, video camera,
refrigerator, computer, microwave oven, washing machine, telephone, automobile,
second home\apartment, piped-in potable water, hot water, sewage system connected
to home.
Regional economic integration in Latin America

Favouring integration (%)
90
80
70
60
50
40
Very bad
Bad
Fair
Good
Opinion of the European Community
Very good
Fig. . Opinion of the European Community and support for integration
in Latin America.
years, however, the EU has supported programmes (seminars, lectures,
etc.) throughout Latin America to popularise their model of economic
integration."* While it is not possible to show a direct linkage between
those efforts and public opinion, the data do show a strong relationship
between a favourable opinion of the EU and support for Latin American
regional integration. Latin Americans also gave their opinion of the USA
in the same question series, a country which came out significantly below
the EU ;  per cent of the respondents gave a positive opinion of the EU
compared to  per cent for the USA. Figure  shows the relationship
between opinion of the EU and support for Latin American economic
integration. Among Latin Americans who have a very negative opinion
of the EU, only one-third support economic integration in Latin America,
whereas among those who have a favourable opinion, over  per cent
support integration.
The close linkage between opinion of the EU and support for Latin
American integration is not merely a generalised support for foreign
countries. An analysis of opinions on the USA and support for integration
does reveal a positive association, but one that is extremely weak and does
not confirm to the linear pattern shown for the EU. Of course, there is
always some risk involved in imitation. If the EU should falter as a trading
unit, Latin Americans might understandably become more sceptical of the
"* For a review of this literature see Simon Francoise and Susan Kaufman Purcell, ‘ The
Impact of Regional Integration on European–Latin American Relations, ’ in Simon
Francoise and Susan Kaufman Purcell (eds.), Europe and Latin America in the World
Economy (Boulder, Colorado, ).

Mitchell A. Seligson
Favouring integration (%)
90
80
70
60
50
Never
Almost never
Frequently
Follow political news
Very frequently
Fig. . Attention to political news and support for integration : by Gender :
——, male ; - - -, female.
model, but the way things stand at the moment, the European model
seems to be responsible for helping to inspire support for Latin American
integration.
Support for democracy
As noted at the outset of this paper, Latin America is institutionally
democratic at the moment, a significant shift from prior decades. It is also
the case that Latin America is moving toward economic integration in
firmer and deeper ways than in the past. Is there any association between
the two ? That is, can it be that support for democracy may translate into
support for an economic model that involves regional integration ? The
survey data suggest an affirmative answer to this question.
The Latinbarometer contains an item that has frequently been used to
measure support for democracy. It reads : ‘ In general, would you say that
you are very satisfied, satisfied, not very satisfied or not at all satisfied with
the way democracy works [fill in country] ? In Latin America ’, there is a
statistically significant ( .) relationship between satisfaction with
democracy and support for economic integration. Support for integration
increases from about two-thirds of the population among those with low
satisfaction with democracy to nearly  per cent for those who are
satisfied with it. Another item in the survey asks : ‘ Would you be willing
to defend democracy if it seemed to be threatened ? ’ Here again the
relationship is statistically significant ( .), with three quarters of
those willing to defend democracy supportive of integration compared to
two-thirds of those among those unwilling to defend it.
Regional economic integration in Latin America

Political attentiveness
The survey provides a number of questions to measure respondents’
political attentiveness. Analysis of each of these reveals the same pattern :
those who are more attentive to politics are more likely to be supporters
of integration. For example, the survey asks about listening to political
news on the TV and radio, and about reading political news in the
newspaper. It also asks a more general question about the extent to which
the respondent ‘ follows political news ’. The pattern for each of these
measures of political attentiveness is similar. In Figure  the results for
the most general question, following political news, are displayed. The
pattern for men is strong and linear, such that fewer than  per cent of
those who never pay attention to political news support integration,
compared to over  per cent who pay frequent attention. Among
females, however, the pattern is far less marked, and even shows a decline
in the group which is the most attentive.
Overall model of support for economic integration
Thus far the analysis has taken a bivariate approach so that the reader
could easily visualize the relationships between each predictor and
support for economic integration in Latin America. Since the predictors
themselves are interrelated (e.g. education is positively associated with a
number of the variables discussed above), it is important to examine the
overall relationship using multivariate techniques. The results presented
here incorporate each of the variables discussed in this paper, employing
as the dependent variable support for integration in Latin America.#!
Table  contains the OLS regression results.
Each of the variables discussed in this paper (except wealth, as already
noted) is a statistically significant predicator of support for integration in
Latin America. A key predictor is perceived benefits of integration ; the
larger the perceived benefits, the greater the support. Also closely related
is the positive association between Latin Americans’ opinion of the
European Union and support for integration. The two key socioeconomic and demographic variables of education and gender are also
significant, with the t-value of education higher than any of the remaining
variables in the model. Gender, however, has a very low t-value when
compared to education. Support for democracy (in the form of willingness
#! The Latinbarometer contains a large number of variables, and those that seemed
theoretically appropriate were examined for this linkages to support for integration.
However, this paper reports on those that remained significant in a multivariate
analysis. Index construction of some of the variables (for example grouping the large
number of questions related to media attentiveness) produce slightly different results,
but the substance remains the same.

Mitchell A. Seligson
Table . OLS Regression : Predictors of Support for Latin American Integration
B
Constant
Benefits of economic integration
Opinion of European Union
Economic situation
Education
Willingness to defend democracy
Attentiveness to political news
Satisfaction with functioning of
democracy
Business owner
Student
Salaried professional
Unemployed
Independent professional
Employee
Gender
Adjusted R# l ., P
.
Beta
t
P
k.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
—
.
.
.
.
.
.
.
k.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
k.
k.
k.
k.
.
k.
k.
.
.
.
.
.
.
k.
k.
k.
k.
k.
.
k.
k.
k.
.
k.
k.
.
k.
k.
.
.
.
.
.
.
.
.. Gender  l male ;  l female.
to defend it and satisfaction with its functioning) also plays an independent
role in predicting support for integration. Those who are more attentive
to political news are more supportive as well. The occupation variable was
divided into a series of dummy variables, so the effect of each of the major
occupational groupings could be seen. Farmers, who were shown in
Figure  to have the lowest support, were used as the base category,
against which the others were compared. Unemployed and student status
were also included as separate categories. As can be seen from the signs
of the coefficients in Table , only the status of independent professional
had a positive association with support for integration (the categories of
high-level executives and mid-level executives, which also had positive
links to support, were not significant once other variables were entered
into the regression analysis). On the other hand, the status of business
owner, student, salaried professional, the unemployed and employee, each
had a negative impact on support. Overall, occupation played only a small
role in determining support as shown by the t-values.
Implications and conclusions
In many ways this paper presents a picture of popular attitudes toward
economic integration in Latin America that would cheer those who
favour further integration. In all but four countries, two-thirds of those
interviewed express support – and in nine out of the  countries, threequarters or more of those with an opinion are in favour of – integration.
Regional economic integration in Latin America

In every country, majorities of opinion holders support integration. Of
course, these findings could be gleaned from looking merely at the
percentages reported in the Latinbarometer results. But the payoff in this
analysis has been the bi-variate and multivariate analysis.
Two economic variables have proven to be directly linked to support
for Latin American integration : perceived benefits of regional trading
blocks and perception of personal and national economic situation. This
finding suggests that success breeds success, and the more that Latin
America prospers from regional trade and the more that its economies
flourish, the greater support there will be for economic integration.
Latin Americans also pay attention to Europe, and the integration
process underway there. The higher the opinion of European integration,
the greater the support for Latin American integration. This finding
suggests that the efforts of the EU and its member nations to promote the
integrationist ideals beyond their shores may be having a positive effect.
Given that no reasonable person can expect completely smooth sailing for
the European process in the years to come, it will be all the more
important for the EU to continue to promote its model in Latin America
so that public opinion can overcome the inevitable disappointments.
Women are less supportive of integration than men in a number of
countries in Latin America. A large part of that difference, however, is a
function of the lower level of education of females as compared to males.
Once education is controlled for, the gender gap in opinion narrows
considerably as the regression analysis shows. Perhaps more important
from a policy perspective is that the highly educated and news-attentive
are strong supporters of integration. Since the highly educated are usually
the most politically active in any society, the ones who vote the most, who
make the most demands on their elected officials, it is likely that popular
support for integration will be especially effective in influencing policy.
Furthermore, the opinions of the well-educated may have a restraining
effect on populist regimes that, from time-to-time, have been elected in
Latin America on the basis of nationalistic appeals. If such regimes seek
to isolate their countries from regional economic integration, they may
well meet with majoritarian popular opposition led by the strongest
opposition from the most highly educated.
Finally, it is gratifying to see that democracy and support for
integration are linked. The more satisfied Latin Americans are with
democracy and the more willing they are to defend it, the greater their
support for economic integration. If Latin America continues along its
recent path of increasing democratisation these findings imply that it will
likely see growing support for economic integration. One might even
speculate that support for economic integration could eventually spill-

Mitchell A. Seligson
over into support for political integration, but such hopes have been
raised before by early theorists of economic integration and have not been
strongly supported.#" Nationalism in Latin America has frequently been
underestimated because of the relative infrequency of international wars
in the region during the present century. Perhaps nationalism will fade
along with the militarism and dictatorship of the past, but the present data
set does not allow the exploration of that possibility. For the moment,
supporters of integration in Latin America will have to content themselves
with the optimistic picture presented here without leaping to the
unjustified conclusion that we are soon to see the emergence of a ‘ United
States of Latin America ’.
#" An early study that showed the sharp discontinuity of economic and political
integration was Mitchell A. Seligson, ‘ Transactions and Community Formation :
Fifteen Years of Growth and Stagnation in Central America ’, Journal of Common Market
Studies, vol.  (March ), pp. –.